Construction and Building Materials 124 (2016) 961–966

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier .com/locate /conbui ldmat

Technical note

Synthesizing one-part geopolymers from rice husk ash

http://dx.doi.org/10.1016/j.conbuildmat.2016.08.0170950-0618/� 2016 Elsevier Ltd. All rights reserved.

⇑ Corresponding author.E-mail address: gregor.gluth@bam.de (G.J.G. Gluth).

P. Sturm a, G.J.G. Gluth a,⇑, H.J.H. Brouwers b, H.-C. Kühne a

aDivision 7.4 Technology of Construction Materials, Bundesanstalt für Materialforschung und -prüfung (BAM), Berlin, GermanybDepartment of the Built Environment, Eindhoven University of Technology, Eindhoven, The Netherlands

a r t i c l e i n f o

Article history:Received 20 June 2016Received in revised form 5 August 2016Accepted 7 August 2016Available online 11 August 2016

Keywords:Alkali-activationGeopolymersOne-part formulationBio-based materialsRice husk ash

a b s t r a c t

One-part geopolymers offer advantages over conventional geopolymers with regard to handling and stor-age of feedstocks. However, they often suffer from a low degree of reaction, a high amount of crystallinebyproducts, and consequently low strength. In this study, one-part geopolymers were produced from ricehusk ash (RHA) and sodium aluminate, and investigated by XRD, ATR-FTIR, SEM and compressivestrength testing. The compressive strength of the material was �30 MPa, i.e. significantly higher thanfor comparable one-part geopolymers. This is attributed to an almost complete reaction of the RHAand the absence of crystalline byproducts (zeolites) in the hardened geopolymer.

� 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Because of their versatile potential, conventional ‘‘two-part”geopolymers, synthesized by activation of solid aluminosilicatefeedstocks (e.g., metakaolin, fly ash and blast-furnace slag) withhighly alkaline liquids (e.g., sodium hydroxide and sodium silicatesolutions), have been subject of a large number of studies in thelast decades (e.g. [1–10]). More recent approaches have focusedon so called ‘‘one-part” formulations. For this route the alkalineactivator is provided in solid form and just water has to be addedto initiate the geopolymerization reactions and hardening(‘‘just add water geopolymers”) [11,12]. In these formulations theactivator solution forms directly in the reaction system, thus thesynthesis procedure is comparable to that for the hydration of ordi-nary Portland cement (OPC)-based binder systems. With thisapproach handling of highly alkaline solutions and problems withaging of these solutions are avoided, which can provide a bettersocial and economic acceptance [12,13]. However, the investigatedone-part systems are often dominated by crystalline tectosilicates(zeolites) and the reported compressive strengths are usuallylower than for conventional two-part geopolymers [11,12,14–19].

Rice husk ash (RHA) is a silica-rich agriculture waste material,typically consisting of 90–95 wt.% amorphous SiO2 [20]. In 2010over 700 million tons of rice were produced worldwide [21] in acontext of a steadily increasing demand. This translates to a pro-duction of about 140 million tons of husk in 2010. Calcination of

the husks leaves about 25 wt.% of the initial mass as ash, the so-called RHA [22]. Highly reactive, amorphous RHAs can be producedat calcination temperatures up to 800 �C [23–26], which is signifi-cantly below the temperatures for the production of conventionalOPC clinkers. Heating at higher temperatures is not beneficialbecause it leads to the formation of crystalline cristobalite and try-dimite in the ash [24], which decreases its reactivity. RHA has beeninvestigated for its use in various industrial applications includingthe use as supplementary cementitious material for conventionalcement based binder systems (e.g. [27–29]). Recent studies inves-tigated the production of sodium silicate solutions from RHA foruse as activating solution for conventional ‘‘two part” geopolymers[30–33].

In this communication we report on the use of RHA as solid pre-cursor in the production of one-part geopolymers. The resultsshow that the RHA is highly reactive and is able to form a typicalgeopolymer gel with higher compressive strength than is generallyobserved for one-part geopolymers.

2. Materials and methods

2.1. Materials

The starting materials were characterized by inductivelycoupled plasma optical emission spectroscopy (ICP-OES), X-raydiffraction (XRD), attenuated total reflection Fourier transforminfrared spectroscopy (ATR FT-IR) and scanning electron

Fig. 2. SEM micrograph of the rice husk ash.

Table 1Chemical composition of the starting materials.

Component RHA NaAlO2

(wt.%) (wt.%)

SiO2 88.49 <0.01Al2O3 0.58 59.74Fe2O3 0.31 0.02TiO2 0.03 <0.01CaO 1.00 0.39MgO 0.88 0.01Na2O 0.24 36.35K2O 2.91 0.01SO3 0.54 0.04P2O5 1.83 n.d.LOI 2.48 2.63

n.d.: not determined; LOI: loss on ignition at 1000 �C.

962 P. Sturm et al. / Construction and Building Materials 124 (2016) 961–966

microscopy (SEM). The experimental conditions are described inSection 2.3.

The employed rice husks from Tanzania were calcined at 650 �Cfor 90 min to yield the rice husk ash (RHA) and the ash thenallowed to cool down naturally to room temperature. After cooling,the RHA was ground in a disc mill for 18 s. Particles that did notpass a sieve with 63 lm mesh size were ground for further 18 s;after the second milling all particles passed the applied sieve.

The particle size distribution (PSD) of the RHA was investigatedusing laser granulometry (Fig. 1). The median of the PSD was deter-mined to be d50 = 11.1 lm (and d90 = 39.8 lm), which is in linewith SEM micrographs of the RHA (Fig. 2). The BET specific surfacearea of the RHA was determined to be 49.6 m2/g (average porediameter �12.9 nm), using N2 sorption at 77 K. An apparent parti-cle density of 2.259 ± 0.003 g/cm3 was determined by Hepycnometry.

The RHA contained 88.49 wt.% SiO2 (Table 1). XRD showed amajor amorphous hump, centered at �22� 2h, and minor amountsof the following crystalline impurities: quartz (PDF # 00-046-1045), potassium magnesium phosphate (PDF # 00-050-0146),cristobalite (PDF # 00-039-1425), and calcite (PDF # 01-086-0174) (Fig. 3).

The employed sodium aluminate (NaAlO2) had an almost stoi-chiometric Na/Al ratio of nearly unity (1.001 mol/mol; cf. Table 1).Besides anhydrous sodium aluminate (PDF # 00-033-1220), XRDalso indicated minor amounts of hydrated sodium aluminate(PDF # 01-083-0315) and natrite (PDF # 01-072-0628). This is inagreement with the low loss on ignition of 2.63 wt.% (Table 1)and the tendency of alkaline compounds to carbonate.

2.2. Geopolymer synthesis

One-part geopolymers were synthesized by mixing RHA andsolid sodium aluminate, and subsequently adding water at a nom-inal water/binder ratio (w/b) of 0.5 by mass to yield molar Na2O:Al2O3:SiO2:H2O ratios as presented in Table 2. The alkaline activa-tor was provided as the sodium aluminate, which dissolves readilyin water, thus water only had to be added to initiate the reactions.The mix-design was chosen to facilitate comparison with previ-ously investigated formulations with the same molar ratios[16,18,19].

Pastes were mixed for 4 min using a contact free planetary cen-trifugal mixer at a rotation speed of 1200 min�1. After mixing thepastes were immediately cast into 20 mm � 20 mm � 20 mm cubemoulds. The samples were cured in the open moulds at 80 �C and

0

10

20

30

40

50

60

70

80

90

100

0.1 1.0 10.0 100.0 1000.0

Volu

me

pass

ing

size

(%)

Particle size (µm)

Fig. 1. Particle size distribution of the rice husk ash.

Fig. 3. XRD patterns of the RHA feedstock and the one-part geopolymers after 1 d,3 d and 7 d of curing. (q = quartz, c = cristobalite, P = potassium magnesiumphosphate; C = calcite, T = thermonatrite.).

80% r.H. in an oven with climate (r.H.) conditioning. The curingtemperature of 80 �C was chosen to accelerate geopolymerizationreactions and hardening, as is standard in geopolymer technology(e.g. [4,5,12,18,30]), though geopolymers from highly reactivefeedstocks in an optimized mix can also be cured at room

Table 2Mix-design and molar ratios of the one-part geopolymers.

m(Na2O) m(Al2O3) m(SiO2) m(H2O) Na2O/Al2O3 SiO2/Al2O3 H2O/Al2O3

(wt.%) (wt.%) (wt.%) (wt.%) (mol/mol) (mol/mol) (mol/mol)

10.17 16.76 34.46 35.03 1.00 3.48 11.81

P. Sturm et al. / Construction and Building Materials 124 (2016) 961–966 963

temperature in a reasonable time [6,9,10,15]. After the desired cur-ing times (24 h, 3 d or 7 d) the specimens were removed from theoven and the moulds, allowed to cool down to room temperature,and tested according to the procedures described in Section 2.3.Curing times longer than 24 h are unusual in geopolymer technol-ogy and would require higher energy consumption; nevertheless,longer curing times were investigated to clarify whether thesecan improve the mechanical properties of the material.

2.3. Structural and mechanical investigations

XRD investigations were conducted on powdered samples on aRigaku Ultima IV device using the following measurement condi-tions: Bragg-Brentano geometry; CuKa radiation (k = 1.5419 Å);scanning range: 5–65� 2h; step size: 0.02� 2h; scanning speed:0.5� 2hmin�1; sample rotation speed: 15 rpm.

ATR FT-IR experiments were performed on powdered sampleson a Bruker Tensor 27 device with a diamond ATR module. Spectrawere recorded in the range 4000–400 cm�1 with a resolution of4 cm�1 and 16 scans per spectrum (detector: RT-DLaTGS).

SEM observations were carried out on a Carl Zeiss EVO MA10device at an accelerating voltage of 7–10 kV using a secondaryelectron (SE) detector. The investigations were conducted on frac-ture surfaces of the specimens after compressive strength testing.Samples were sputtered with gold before the SEM measurements.

The compressive strength tests were performed on a ToniPRAXdevice at a loading rate of 240 N/s, which is equivalent to 0.6 MPa/sfor the investigated specimens (cubes, 20 mm � 20 mm � 20 mm).Testing was always conducted within the first 30 min after remov-ing the samples from the curing regime. Compressive strength wastested after 1 d, 3 d and 7 d of curing.

Fig. 4. ATR FT-IR spectra of the RHA feedstock and the one-part geopolymers after1 d, 3 d and 7 d of curing. (1 = mas,s H2O; 2 = d H2O; 3 = carbonate; 4 = mas Si-O-Si;5 = mas Si-O-T (T = Si or Al); 6 = ms Si-O-Si; 7 = ms Si-O-T (T = Si or Al); 8 = CB (complexband); 9 = d O-Si-O in RHA; 10 = d O-Si-O in geopolymers.)

3. Results and discussion

3.1. Structural investigations

The XRD results for the untreated RHA as well as for the curedgeopolymers are shown in Fig. 3. The hump from the RHA, maxi-mum located at 21.82� 2h, disappeared after curing and a newhump with its maximum at 27.68� 2h, representing the geopoly-meric gel [34], occurred after curing. The disappearance of thehump from the RHA indicates its virtually complete reaction, i.e.a degree of reaction of the RHA close to 100%. This contrasts previ-ously investigated one-part geopolymer systems, in which thesilica feedstocks reacted only partly, depending on the initial SiO2/Al2O3 ratio [16,18]. In particular, at a starting SiO2/Al2O3 (i.e. totalSiO2/Al2O3 ratio of the mix including silica and sodium aluminate)of �3.5 mol/mol the degree of reaction of the silica feedstocks(microsilica or silica-rich residue from chlorosilane production)in these previous investigations was only �55%, while the degreeof reaction of the RHA at the same starting SiO2/Al2O3 (Table 2)is near 100%.

The diffractograms of the 1 d, 3 d and 7 d cured specimens indi-cated almost completely amorphous reaction products. No crys-talline aluminosilicates, such as zeolites, were observed; thus thecured mixes can be considered to be pure geopolymers, different

to the composite-like character of zeolite-containing one-part sys-tems mentioned above [16,18,19].

Minor amounts of thermonatrite (Na2CO3�H2O; PDF # 00-008-0448) could be identified in the geopolymers, due to minor carbon-ation of the highly alkaline reaction system. The phosphate and thecalcite, present in the starting RHA, disappeared after curing, indi-cating their dissolution during curing; cristobalite and quartzremained in the system after curing (Fig. 3).

Curing for longer periods than 1 d did not lead to significantchanges in the phase content, as observed by XRD. This leads tothe conclusion that after one day of curing the geopolymerizationreactions were almost complete. This behavior has also beenobserved with one-part geopolymers based on other silica feed-stocks cured at 80 �C [16,18].

The infrared (ATR FT-IR) spectra of the unreacted RHA and thecured geopolymers are shown in Fig. 4. All spectra, particularlythe spectra of the cured geopolymers, exhibited broad and overlap-ping absorption bands, which demonstrates the amorphous char-acter of the compounds.

Bands around 3350 cm�1 (vibration 1; numbers refer to desig-nations in Fig. 4) correspond to overlapping asymmetric and sym-metric stretching vibrations of the incorporated water and are onlypresent in the geopolymer samples. Their intensity decreased withincreasing curing time, indicating a slight water loss during curing,despite the fact that relative humidity was 80%. The same behavior– a decrease in intensity with increasing curing time – wasobserved for the bending vibration of water (vibration 2). No waterbands were observed in the RHA spectrum, consistent with the factthat the material was produced by calcination at 650 �C.

Bands around 1440–1400 cm�1 (vibration 3) in the geopoly-mers are attributed to carbonate [35], i.e. the signals are causedby the incorporated thermonatrite, which has been observed byXRD (Fig. 3). These carbonate absorptions bands were only foundin the geopolymer samples, i.e. the marginal amount of calcite inthe anhydrous RHA was not observed in its infrared spectrum.

Peaks at 1059 cm�1 (vibration 4) and between 985 and975 cm�1 (vibration 5) correspond to asymmetric stretching vibra-tions of Si-O-Si in the RHA and Si-O-T (T = Si or Al) in the geopoly-mers, respectively. The absorption bands of asymmetric Si-O-T

964 P. Sturm et al. / Construction and Building Materials 124 (2016) 961–966

stretching vibrations in tectosilicates shift to lower wavenumberswith increasing aluminum content [36–38]. Therefore, this shiftproves the incorporation of aluminum in the geopolymer network.

The Si-O-Si symmetric stretching vibration of the RHA at797 cm�1 (vibration 6) did not remain in the geopolymer samples.Instead, Si-O-T (T = Si or Al) absorptions were observed at 684–680 cm�1 (vibration 7). Again, there was a shift to lower wavenum-bers and the disappearance of the characteristic vibration 6(ms Si-O-Si) was a further indication of a high degree of reactionof the silica feedstock.

Absorption bands in the range 575–573 cm�1 (vibration 8) inthe geopolymers are attributed to complex vibrations (CB =complex band) of stretching and bending vibrations (ms Si-O-Si; dO-Si-O) [39]. Bands at 451 cm�1 (vibration 9) and at 421–419 cm�1

(vibration 10) represent O-Si-O bending vibrations in the RHAfeedstock and the geopolymers, respectively.

Fig. 5. SEM micrographs of the one-part geopolymers after 3 d of curing.

SEM micrographs of the 3 d cured one-part geopolymer areshown in Fig. 5. At lower magnification (Fig. 5, top), a microstruc-ture comprising homogeneously distributed, small pores, wasobserved. At higher magnification (Fig. 5, bottom), the geopolymerexhibits mainly the dense and homogeneous microstructure that isobserved for well-cured conventional geopolymers (e.g., based onfly ash [40]) too. Besides, the micrograph also shows regions of par-ticulate structure (e.g., bottom middle); these particles are how-ever not remnants of the RHA, as the latter has reacted almostcompletely.

3.2. Compressive strength

Fig. 6 shows the compressive strengths of the one-part geopoly-mer after 1, 3 and 7 days of curing. After 1 d the specimens reachedcompressive strengths of 29.8 MPa on average. After 3 d of curingthe average compressive strength reached the maximum value of32.7 MPa. Curing for 7 d resulted in a slight strength decrease(30.1 MPa). Since the compressive strengths at all investigated cur-ing times were similar, it is concluded that curing times longerthan 24 h are of no use for the investigated material at theemployed curing conditions.

The obtained strengths are significantly higher than thestrengths of one-part geopolymers with similar compositions, butdifferent silica starting materials [16,18,19]. The principal reasonfor the higher strength of the one-part geopolymers produced fromRHA is assumed to be the higher degree of reaction of the silicaand the absence of zeolites, as is discussed in more detail inSection 3.3.

3.3. General discussion

Generally, one-part geopolymer mixes tend to form significantamounts of crystalline zeolites as reaction products [11,12,14–19]. In addition, it appears that these mixes often yield productsof comparatively low compressive strengths; in the published liter-ature strengths are significantly below 30 MPa or strength data isnot reported [11,12,14,16–19]. However, Feng et al. [15] were ableto produce one-part geopolymers with a comparatively lowamount of incorporated crystallites, which reached compressivestrengths of >30 MPa after 7 days of curing at 25 �C. This suggeststhat avoiding the formation of crystalline byproducts duringgeopolymerization enhances the strengths of one-part geopolymers.

0

5

10

15

20

25

30

35

40

0 1 2 3 4 5 6 7 8

Com

pres

sive

str

engt

h (M

Pa)

Curing time (d)

Fig. 6. Compressive strength of the one-part geopolymers after 1 d, 3 d and 7 d ofcuring. (Error bars represent one standard deviation in each direction.).

P. Sturm et al. / Construction and Building Materials 124 (2016) 961–966 965

It is thus interpreted that the absence of crystalline side-products(zeolites) in the RHA-based geopolymers (Fig. 3) contributed totheir higher strength when compared to one-part geopolymersproduced from most other feedstocks.

The one-part mixes investigated in Refs. [16,18,19], synthesizedin the same way as the formulations investigated in the presentcontribution, but using microsilica and a residue from chlorosilaneproduction as silica feedstocks, were found to have an silica/alumina-ratio of the reactions products of SiO2/Al2O3 � 2 mol/mol,independent of the starting mix-design of the pastes [16,18]. In allcases, dissolution of the silica and formation of reaction productsceased when this SiO2/Al2O3 ratio was reached. The silica not con-sumed to reach this SiO2/Al2O3 ratio remained undissolved in thehardened pastes. In contrast to these previous results, the one-part geopolymers produced from RHA exhibited a virtually com-plete reaction of the silica feedstock (RHA), as indicated by XRDand ATR FT-IR spectroscopy, and the pastes transformed intoamorphous sodium aluminosilicate (N-A-S(-H)) gel, i.e. no zeolitesformed. It is known that it is this N-A-S(-H) gel that is mainlyresponsible for the strength development of low-calcium alkali-activated materials [2,34]. The strength of geopolymers is thusdependent on the amount of N-A-S(-H) gel formed, its compositionand its microstructure. As shown in Fig. 5 (bottom) the RHA-basedone-part geopolymers exhibited a rather dense and homogeneousmicrostructure. In contrast, the microstructure of the one-partgeopolymers based on the other silica materials (i.e. microsilicaor silica residue from chlorosilane production) consists of ca.100 nm-sized particles, possessing a large amount of interparticlepore space [16,19]. This difference can be attributed to theSiO2/Al2O3 ratio of the geopolymeric gel: In their SEM study onmetakaolin- and fly ash-based geopolymers, Steveson andSagoe-Crentsil [41,42] found that there is a transition from aparticulate and porous microstructure at low SiO2/Al2O3 ratios(2.5–3 mol/mol) to a dense and homogeneous microstructure athigher SiO2/Al2O3 ratios (3.5–3.9 mol/mol). Similar observationshave been made by Duxson et al. [43]. It can thus be concluded thatthe high degree of reaction of the RHA led to a comparatively highSiO2/Al2O3 ratio of the reacted geopolymer (presumably inthe range of the starting SiO2/Al2O3 of �3.5 mol/mol; Table 2) andconsequently to a denser microstructure than in other one-partgeopolymers, which caused a higher compressive strength of theRHA-based material.

Summarizing, we attribute the comparatively high compressivestrengths of the RHA-based one-part geopolymers mainly to (1)the absence of significant amounts of crystalline byproducts, and(2) the almost complete reaction of the RHA, which resulted in ahigher SiO2/Al2O3 ratio and a denser microstructure of the N-A-S(-H) gel than in one-part geopolymers produced from other silicasources.

Other factors that probably also contributed to the compara-tively high compressive strength of the RHA-based one-partgeopolymers were: (1) the absence of swelling of the hardeningpaste, contrary to observations made on some one-part geopoly-mers based on microsilica or other silica-rich by-products [19],and (2) a better workability of the RHA-based pastes as comparedto pastes based on the other aforementioned silica feedstocks,which can lead to less inhomogeneities in the paste and thus inthe hardened geopolymer. The latter observation is attributed todifferent particle characteristics of the RHA: The particle size dis-tribution as measured by laser granulometry (d50 = 11.1 lm;d90 = 39.8 lm; Fig. 1) of the RHA was significantly coarser thanthe particle size distributions of previously investigated silica feed-stocks – microsilica and a residue from chlorosilane production,which had d50 � 0.2–1.0 lm and 6.8 lm, respectively [16,18]. Inaddition, these other silica feedstocks consisted of particles com-posed of small (d � 100–200 nm) primary particles, causing a high

intraparticle porosity, which increases the water demand of thepastes and leads to a poorer workability.

Despite its coarser PSD, the RHA had a higher specific surfacearea (49.6 m2/g) than the microsilica and the residue fromchlorosilane production (20.1 m2/g and 32.3 m2/g, respectively).This can be explained by the fact that the particles of the RHAhad a more angular shape, apparently a rougher surface (cf.Fig. 2) and a significant amount of mesopores (average pore diam-eter �12.9 nm; see Section 2.1) when compared to the particles ofmicrosilica and similar materials. This higher specific surface areaof the RHA is assumed to be a major reason for the higher degree ofreaction of the RHA when compared to the other silica feedstocks.

4. Conclusions

One-part geopolymers were produced from rice husk ash (RHA)by mixing the RHA with solid sodium aluminate at a ratio to yieldSiO2/Al2O3 = 3.5 and subsequently mixing the solid blend withwater and curing at 80 �C and 80% r.H. The compressive strengthof the geopolymer was �30 MPa after one day of curing. As shownby XRD, the feedstocks almost completely reacted and transformedinto amorphous geopolymer (N-A-S(-H) gel). Comparison of theinfrared spectra of the RHA and the reaction products indicated ahigh degree of reaction of the RHA too. No crystalline reactionproducts were observed, except for minor amounts of hydratedsodium carbonate due to carbonation. Curing for longer times thanone day caused a slight decrease of the water content of thegeopolymer, but no other significant changes were observed inthe XRD patterns and the infrared spectra.

The results of this study thus show that RHA can be appliedsuccessfully for the synthesis of one-part geopolymers. TheRHA-based geopolymers provide some improved properties whencompared to one-part geopolymers based on other silica materials,in particular an almost complete reaction of the silica feedstock anda comparatively high compressive strength. Optimization of theRHA-based one-part geopolymers regarding chemical compositionand water/binder ratio, as well as curing conditions and curingtimes will be subject of future studies.

Acknowledgements

The authors thank Christiane Weimann and Carsten Prinz forhelp with the SEM investigations and the determination of specificsurface area and porosity of the RHA, respectively. The authors arealso very grateful to Nsesheye Msinjili for providing the rice huskash sample. P.S. acknowledges financial support from BAM withinthe MIS program (proposal Ideen_2012_36).

References

[1] P. Duxson, A. Fernández-Jiménez, J.L. Provis, G.C. Lukey, A. Palomo, J.S.J. vanDeventer, Geopolymer technology: the current state of the art, J. Mater. Sci. 42(2007) 2917–2933.

[2] J.L. Provis, S.A. Bernal, Geopolymers and related alkali-activated materials,Annu. Rev. Mater. Res. 44 (2014) 299–327.

[3] S.A. Bernal, E.D. Rodríguez, A.P. Kirchheim, J.L. Provis, Management andvalorisation of wastes through use in producing alkali-activated cementmaterials, J. Chem. Technol. Biotechnol. 91 (2016) 2365–2388.

[4] C. Gunasekara, D.W. Law, S. Setunge, Long term permeation properties ofdifferent fly ash geopolymer concretes, Constr. Build. Mater. 124 (2016) 352–362.

[5] M. Babaee, A. Castel, Chloride-induced corrosion of reinforcement in low-calcium fly ash-based geopolymer concrete, Cem. Concr. Res. 88 (2016) 96–107.

[6] X. Gao, Q.L. Yu, H.J.H. Brouwers, Reaction kinetics, gel character and strength ofambient temperature cured alkali activated slag-fly ash blends, Constr. Build.Mater. 80 (2015) 105–115.

[7] K. Dombrowski, M. Weil, A. Buchwald (Geopolymere Bindemittel. Teil 2:Entwicklung und Optimierung von Geopolymerbetonmischungen für feste unddauerhafte Außenwandbauteile), ZKG Int. 61 (3) (2008) 70–80.

966 P. Sturm et al. / Construction and Building Materials 124 (2016) 961–966

[8] R. Pouhet, M. Cyr, Alkali-silica reaction in metakaolin-based geopolymermortar, Mater. Struct. 48 (2015) 571–583.

[9] M.C. Bignozzi, S. Manzi, M.E. Natali, W.D.A. Rickard, A. van Riessen, Roomtemperature alkali activation of fly ash: effect of Na2O/SiO2 ratio, Constr. Build.Mater. 69 (2014) 262–270.

[10] W.D.A. Rickard, G.J.G. Gluth, K. Pistol, In-situ thermo-mechanical testing of flyash geopolymer concretes made with quartz and expanded clay aggregates,Cem. Concr. Res. 80 (2016) 33–43.

[11] D. Koloušek, J. Brus, M. Urbanova, J. Andertova, V. Hulinsky, J. Vorel,Preparation, structure and hydrothermal stability of alternative (sodiumsilicate-free) geopolymers, J. Mater. Sci. 42 (2007) 9267–9275.

[12] A. Hajimohammadi, J.L. Provis, J.S.J. van Deventer, One-part geopolymer mixesfrom geothermal silica and sodium aluminate, Ind. Eng. Chem. Res. 47 (2008)9396–9405.

[13] P. Duxson, J.L. Provis, Designing precursors for geopolymer cements, J. Am.Ceram. Soc. 91 (2008) 3864–3869.

[14] S.J. O’Connor, K.J.D. MacKenzie, Synthesis, characterisation and thermalbehaviour of lithium aluminosilicate inorganic polymers, J. Mater. Sci. 45(2010) 3707–3713.

[15] D. Feng, J.L. Provis, J.S.J. van Deventer, Thermal activation of albite for thesynthesis of one-part mix geopolymers, J. Am. Ceram. Soc. 95 (2012) 565–572.

[16] G.J.G. Gluth, C. Lehmann, K. Rübner, H.-C. Kühne, Geopolymerization of a silicaresidue from waste treatment of chlorosilane production, Mater. Struct. 46(2013) 1291–1298.

[17] X. Ke, S.A. Bernal, N. Ye, J.L. Provis, J. Yang, One-part geopolymers based onthermally treated red mud/NaOH blends, J. Am. Ceram. Soc. 98 (2015) 5–11.

[18] P. Sturm, S. Greiser, G.J.G. Gluth, C. Jäger, H.J.H. Brouwers, Degree of reactionand phase content of silica-based one-part geopolymers investigated usingchemical and NMR spectroscopic methods, J. Mater. Sci. 50 (2015) 6768–6778.

[19] P. Sturm, G.J.G. Gluth, S. Simon, H.J.H. Brouwers, H.-C. Kühne, The effect of heattreatment on the mechanical and structural properties of one-partgeopolymer-zeolite composites, Thermochim. Acta 635 (2016) 41–58.

[20] J. He, Synthesis and characterization of geopolymers for infrastructuralapplications (Ph.D. thesis), Louisiana State University, Baton Rouge, 2012.

[21] FAO Statistical Yearbook – World Food and Agriculture, Food and AgricultureOrganization of the United Nations, Rome, 2013, p. 2013.

[22] D. Koteswara Rao, P.R.T. Pranav, M. Anusha, Stabilization of expansive soil withrice husk ash, lime and gypsum – an experimental study, Int. J. Eng. Sci.Technol. 3 (2011) 8076–8085.

[23] S. Chanddrasekhar, K.G. Satyanarayana, P.N. Pramada, Processing, propertiesand applications of reactive silica from rice husk—an overview, J. Mater. Sci. 38(2003) 3159–3168.

[24] Y. Shinohara, N. Kohyama, Quantitative analysis of tridymite and cristobalitecrystallized in rice husk ash by heating, Ind. Health 42 (2004) 277–285.

[25] D.V. Reddy, M. Alvarez, Marine durability characteristics of rice husk ash-modified reinforced concrete, in: International Latin American and CaribbeanConference for Engineering and Technology, Mayagüez, Puerto Rico, 2006.

[26] G. Abood Habeeb, H. Bin Mahmud, Study on properties of rice husk ash and itsuse as cement replacement material, Mater. Res. 13 (2010) 185–190.

[27] G. Rodríguez de Sensale, Effect of rice-husk ash on durability of cementitiousmaterials, Cem. Concr. Compos. 32 (2010) 718–725.

[28] G. Chagas Cordeiro, R. Dias Toledo Filho, L.M. Tavares, E. de Moraes RegoFairbairn, S. Hempel, Influence of particle size and specific surface area on thepozzolanic activity of residual rice husk ash, Cem. Concr. Compos. 33 (2011)529–534.

[29] V.-T.-A. Van, C. Rößler, D.-D. Bui, H.-M. Ludwig, Rice husk ash as bothpozzolanic admixture and internal curing agent in ultra-high performanceconcrete, Cem. Concr. Compos. 53 (2014) 270–278.

[30] S.A. Bernal, E.D. Rodríguez, R. Mejia de Gutiérrez, J.L. Provis, S. Delvasto,Activation of metakaolin/slag blends using alkaline solutions based onchemically modified silica fume and rice husk ash, Waste Biomass Valoriz. 3(2012) 99–108.

[31] J. He, Y. Jie, J. Zhang, Y. Yu, G. Zhang, Synthesis and characterization of red mudand rice husk ash-based geopolymer composites, Cem. Concr. Compos. 37(2013) 108–118.

[32] J.M. Meija, R. Mejia de Gutierrez, F. Puertas, Rice husk ash as a source of silicain alkali-activated fly ash and granulated blast furnace slag systems, Mater.Constr. 63 (2013) 361–375.

[33] U.H. Heo, K. Sankar, W.M. Kriven, S.S. Musil, Rice husk ash as a silica source in ageopolymer formulation, Ceram. Eng. Sci. Proc. 35 (8) (2015) 87–101.

[34] J.L. Provis, G.C. Lukey, J.S.J. van Deventer, Do geopolymers actually containnanocrystalline zeolites? A reexamination of existing results, Chem. Mater. 17(2005) 3075–3085.

[35] W.B. White, The carbonate minerals, in: V.C. Farmer (Ed.), The Infrared Spectraof Minerals, Mineralogical Society, London, 1974, pp. 227–284.

[36] H.H.W. Moenke, Silica, the three-dimensional silicates, borosilicates andberyllium silicates, in: V.C. Farmer (Ed.), The Infrared Spectra of Minerals,Mineralogical Society, London, 1974, pp. 365–382.

[37] S. Jakobbson, Determination of Si/Al ratios in semicrystalline aluminosilicatesby FT-IR spectroscopy, Appl. Spectrosc. 56 (2002) 797–799.

[38] W. Mozgawa, W. Jastrebski, M. Handke, Vibrational spectra of D4R and D6Rstructural units, J. Mol. Struct. 744–747 (2005) 663–670.

[39] W. Mozgawa, M. Król, K. Barczyk, FT-IR studies of zeolites from differentstructural groups, Chemik (2011) 671–674.

[40] W.D.A. Rickard, J. Temuujin, A. van Riessen, Thermal analysis of geopolymerpastes synthesised from five fly ashes of variable composition, J. Non-Cryst.Solids 358 (2012) 1830–1839.

[41] M. Steveson, K. Sagoe-Crentsil, Relationships between composition, structureand strength of inorganic polymers. Part I—Metakaolin-derived inorganicpolymers, J. Mater. Sci. 40 (2005) 2023–2036.

[42] M. Steveson, K. Sagoe-Crentsil, Relationships between composition, structureand strength of inorganic polymers. Part 2—fly ash-derived inorganicpolymers, J. Mater. Sci. 40 (2005) 4247–4259.

[43] P. Duxson, J.L. Provis, G.C. Lukey, S.W. Mallicoat, W.M. Kriven, J.S.J. vanDeventer, Understanding the relationship between geopolymer composition,microstructure and mechanical properties, Colloids Surf. A 269 (2005) 47–58.